The debate about whether robots will one day replace human workers has become even more intense since the launch of new AI software such as ChatGPT. But how has digitization actually affected the labor market so far? Where have jobs disappeared and where have new ones been created?
The Swiss labor market
Although digitization has permanently changed the economy in all industrialized countries, its effect on labor markets varies by country. These differences are mainly driven by countries’ different education systems. Due to the dual education system in Switzerland, more than 50% of its workforce has completed vocational education. Just over 10% of all employees are university graduates, and around 20% have only completed compulsory education. The developments in the Swiss labor market are shown in figure 1.
Digitization and its impact on the Swiss labor market
To investigate the effects of digitization on the Swiss labor market, the two economists Filippo Pusterla
Since vocational education is practically non-existent in the US, its labor market differs fundamentally from the Swiss one. Accordingly, the effects of digitization also differ.
Negative effects of digitization in the US were mostly found in so-called routine jobs. Routine jobs can often easily be automated, but they tend to be performed by mid-skill rather than low skill workers. Typical examples are secretaries and accountants.
Low skill jobs, such as cleaning or waiting on tables, on the other hand, are often not easy to automate. Accordingly, digitization has not had a negative impact on low-skill US workers.
To measure the degree of digitization, the researchers calculate the proportion of employees of a firm who work in occupations related to information and communications technology (ICT). Such occupations include, for example, software developers, web and multimedia developers, and specialists for databases and networks.
In a next step, they investigate how the composition of employees by educational attainment has changed in companies with a higher share of workers in ICT over time. To avoid measuring a simple mechanical increase in the share of university workers driven by the usually highly educated ICT workers, the researchers do not consider employees in ICT occupations when computing changes in the educational composition of a firm’s workforce.
Their results show that companies with a higher share of employees in ICT over time (i.e., with a higher degree of digitization) typically employ fewer workers with vocational training or compulsory schooling, and more workers with a diploma of higher education, a degree from a university of applied sciences, or a university degree. The researchers found the greatest positive effect of digitization for employees with a degree from a university of applied sciences. And although employees with only compulsory education are also negatively affected, the researchers find the greatest negative effect for employees with a vocational education degree.
In summary, the study finds that digitization in Switzerland has led to higher demand for employees with a degree from a university of applied sciences in particular, but also with a diploma of higher education or university degree. The demand for employees with vocational education or for low-skilled workers, on the other hand, has fallen with digitization.
 Filippo Pusterla is a senior lecturer and senior researcher at the Swiss Federal University for Vocational Training and Education (EHB).
 Ursula Renold is a professor at the Swiss Federal Institute of Technology in Zurich (ETH Zürich).
 Source: Pusterla, Filippo & Renold, Ursula (2022). "Does ICT affect the demand for vocationally educated workers?", Swiss Journal of Economics and Statistics, 158(22).
 Source: Autor, David, Levy, Frank, & Murnane, Richard (2003) , "The skill content of recent technological change: An empirical exploration.", The Quarterly Journal of Economics, 118(4), 1279–1333.
 Source: Autor, David, Katz, Lawrence, & Kearney, Melissa (2006), "The polarization of the US labor market.", The American Economic Review, 96(2), 189–194.